PSU for digital load, to choke or not to choke?!

[Gappo]

Hi, I'm in the middle of designing a power supply for an i2s audio reclocker. As this reclocker (allo Kali, 5V 100mA) is the masterclock for my diy DAC, I wanted to give him the cleanest power possible to reduce jitter to a minimum.
My first plan was to go with an R-Core transformer I already own, going directly into the first choke of an LCLC filter, like in the circuit shown below (I would add a bleeder resistor in parallel to the first cap which is not shown in the circuit diagram).
For regulation and some more filtering I already have a sparkos labs SS1117-05 voltage regulator. I plan on putting it with an C_in and an C_out cap very close to the reclocker board (as I guess is also recommended by dave in his video about bypass capacitors).
Given specs for the sparkos voltage regulator:
125dB input rejection
3.2uV RMS Output oise in a 20KhzBandwidth @ 5 Vo
2mV Load Regulation

So now to my question(s): Are chokes generally a bad decision for this scenario, as the they deliver a constant current but my load, as it is a digital device, needs a constant voltage. So can it be that the constant current flow has even a negative effect? Like when the current demand of the load is alternating but the chokes want to deliver a constant current flow? Or are the ldo’s and caps taking care of this problem?
Also are the chokes here an absolute overkill and won’t give ma any benefit over a “bunch of caps”  and / or a capacitance multiplier? (I mean just quality wise not in a quality to price/size ratio)

Thanks a lot for any help!
greetings Gappo

[Kleinstein]

One can use the inductor to reduce ripple and improve the power factor. However at the low power it is not really needed as the transformer usually has quite some resistance. Also 2 chokes are usually not needed, if there is a regulator. Normally one does not even use a single inductor at this power level. There is some advantage of the inductor, even if the inductor has a relatively large resistance. It helps to reduce the ripple current at a relative moderate cost.

[Gappo]

I see, so I’ll guess one choke is a good start, thanks for the info!

I forgot to add a question: As I’m also thinking of buying my first oscilloscope, would something in the range of the Rigol DS1054Z also allow me to analyze the jitter performance (so I can check what changes to the power supply quality actually influence clock performance) or are entry level oscilloscopes not fit for this kind of measurements? (It would also be nice to know at what price tag I would be looking for an osci that can do such measurements. I saw a few with an i2s interpreter but they seemed rather pricey ^^)

[exe]

Depends on how much jitter you want to measure. An oscilloscope itself has jitter too, esp. DS1054Z.

Anyway, measurable != audible. Not that I discourage you from doing your research, but there is a possibility to get yourself trapped into "opamp rolling", capacitor replacement, seeing jitter, special cables, etc.

[Gappo]

Yeah there are quite some traps In the hifi scene for sure

 I know there is a huge discussion around jitter and it’s audibility, my hope was that a decent entry level osci is already capable of measuring into the realm of the disputable jitter second fractions (so I can check what shielding and noise filters etc. are improving the jitter, even if it’s not audible. Just because these would be just further design questions of the dac and it’s psu, with no real costs involved. Proper design is nice to have
 
If I can for example check that there is no difference between a smps and my diy psu i can save some money on the choke, if not I might get one choke for the psu and if this is better than my Psu with filter caps I might get another choke just for the heck of having a really nice jitter free dac build by myself.

Edit: evening lecture now are whitepapers on jitter audibility, thanks
Edit2: just saw the jitter specs of the board I use (typical usb to i2s converters in similar range of product quality have similar given specs) are in the <5ps range. If that’s true I guess no entry level oscilloscope will be able to measure this? (On the pro side it’s probably not necessary either ^^)